Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/loulanyue/spec-kit-zhWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/loulanyue/spec-kit-zh/clarify)<a href="https://agentmods.dev/commands/loulanyue/spec-kit-zh/clarify"><img src="https://agentmods.dev/badge/commands/loulanyue/spec-kit-zh/clarify.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00028 | $0.03037 |
| Opus 5 | $0.00014 | $0.01519 |
| Sonnet 5 | $0.00006 | $0.00607 |
| Haiku 4.5 | $0.00003 | $0.00304 |
Grade A, and why
clarify scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 214 lines — stays where its author put it; the contents beside it link to each section on GitHub.
用户输入
$ARGUMENTS
在继续之前,你必须考虑用户输入(如果不为空)。
语言要求
- 所有澄清问题、建议、表格、规范更新、覆盖摘要和完成报告都必须使用简体中文。
- 为保证精确性,技术术语、ID 和文件路径在需要时保留原始形式。
概述
目标:识别并减少当前功能规范中的歧义与缺失决策点,并将澄清结果直接写回规范文件。
说明:该澄清流程应当在执行计划命令之前完成。大多数 agent 使用 /speckit.plan,Codex CLI 使用 /prompts:speckit-plan。如果用户明确表示跳过澄清(例如仅做探索性验证),则可以继续,但必须提醒其后续返工风险会提高。
执行步骤:
第 1 步:获取前置环境信息
从仓库根目录执行 {SCRIPT} 一次(使用 --json --paths-only / -Json -PathsOnly 模式),解析以下最小 JSON 字段:
FEATURE_DIRFEATURE_SPEC- (可选:捕获
IMPL_PLAN、TASKS以便后续链式流程使用)
若 JSON 解析失败,立即中止并告知用户重新运行 spec 命令(/speckit.specify 或 Codex /prompts:speckit-specify)或检查功能分支环境。
参数中如包含单引号(如 "I'm Groot"),请使用转义语法:
'I'\''m Groot'(或尽量使用双引号:"I'm Groot")。
第 2 步:加载规范并执行结构化歧义扫描
加载当前规范文件,按以下分类体系进行覆盖度扫描。对每个类别标记状态:清晰 / 部分 / 缺失。生成内部覆盖度映射用于优先级排序(除非不会提出任何问题,否则不直接输出原始映射)。
功能范围与行为
- 核心用户目标与成功标准
- 明确的范围外声明
- 用户角色 / 用户画像的区分
领域与数据模型
- 实体、属性、关系
- 标识与唯一性规则
- 生命周期 / 状态转换
- 数据量 / 规模假设
交互与用户体验流程
- 关键用户旅程 / 操作序列
- 错误 / 空状态 / 加载状态
- 无障碍访问或本地化说明
非功能质量属性
- 性能(延迟、吞吐量目标)
- 可扩展性(水平/垂直扩展、容量上限)
- 可靠性与可用性(正常运行时间、恢复预期)
- 可观测性(日志、指标、链路追踪信号)
- 安全与隐私(认证/授权、数据保护、威胁假设)
- 合规 / 监管约束(如有)
集成与外部依赖
- 外部服务 / API 及其故障模式
- 数据导入/导出格式
- 协议 / 版本假设
边界情况与故障处理
- 负向场景
- 限流 / 节流
- 冲突解决(如并发编辑)
约束与权衡
- 技术约束(语言、存储、托管)
- 明确的权衡或已拒绝的替代方案
术语与一致性
- 规范词汇表中的权威术语
- 需避免的同义词 / 已废弃术语
完成信号
- 验收标准可测试性
- 可量化的完成定义(Definition of Done)指标
其他 / 占位符
- TODO 标记 / 未决决策
- 缺乏量化的模糊形容词(如"健壮""直观")
对每个状态为部分或缺失的类别,添加候选问题机会,除非:
- 澄清该问题不会实质性影响实现或验证策略
- 该信息更适合在计划阶段再明确(在内部记录)
第 3 步:生成优先澄清问题队列
在内部生成最多 5 个候选澄清问题(不得一次性全部输出)。须遵守以下约束:
- 整个会话最多 5 个问题。
- 每个问题必须可通过以下方式之一回答:
- 简短的多选(2–5 个互斥选项),或
- 一词 / 短语回答(明确约束:"请用 ≤5 个词回答")。
- 只包含答案会实质性影响架构、数据建模、任务分解、测试设计、用户体验行为、运营就绪度或合规验证的问题。
- 确保类别覆盖均衡:优先覆盖影响最高的未解决类别;避免在单个高影响领域(如安全态势)未解决时询问两个低影响问题。
- 排除已回答的问题、琐碎的风格偏好或计划层面的执行细节(除非会阻塞正确性)。
- 优先选择能降低下游返工风险或防止验收测试不对齐的澄清问题。
- 若 5 个以上类别仍未解决,按"影响 × 不确定性"启发式选取前 5 个。
第 4 步:顺序提问循环(交互式)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 214 lines · 28 tokens per session scan A cc1949b7f5a5
clarify is a command published in the GitHub repository loulanyue/spec-kit-zh (339 stars, last pushed 5d ago), licensed MIT. It adds 28 tokens to every session and 3,037 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
sdd-continue
Continue the next SDD phase in the dependency chain.
sdd-apply
Implement SDD tasks — writes code following specs and design.
sdd-status
Show structured SDD status for an active change.
gentle-sdd-ff
Fast-forward all SDD planning phases — proposal through tasks.
sdd-explore
Explore and investigate an idea or feature — reads codebase and compares approaches.
sdd-init
Initialize SDD context — detects project stack and bootstraps persistence backend.